Director, Learning Design & Development| PMIAsheville, NC, United States
With Generative AI, iteratively refining and optimizing prompts can lead to better AI-generated results. This may involve adjusting the specificity or clarity of the prompt to increase relevance and accuracy of results.
What examples do you have of how improving a prompt drastically changed the output quality? What specific changes did you make that led to the improvement?
This is a fundamental principle of interacting with generative models. Progressive refinement of prompts is a process of correcting and enriching the training signal. The more we refine the query, the better the model understands our vision and the more accurately it generates the answer. Saving Changes...
Luis VargasProject Manager| SLBHouston, United States
The tool has a big potential. Our leading role using this tool can significantly impact the output quality by providing clearer instructions and context for the AI to generate more relevant and accurate responses. That is quite good. Saving Changes...
Luis VargasProject Manager| SLBHouston, United States
The right way to use LLMs to generate a clear and concise answer is to be able to refine and link the prompts appropriately. It is amazing how the relevancy of the outcome improves once the prompts are aligned accordingly. Saving Changes...
ABDULAZEEM FADULcivil engineer| Dar Al Riyadh CompanyAbha - Saudi Arabia, Saudi Arabia
Prompt refinement can transform a response from generic and uninspiring to insightful and actionable. By adjusting for clarity, context, and structure, you leverage GenAI’s full potential to deliver tailored, high-quality outputs that align with your goals. Saving Changes...
Asking for the Risks list brought by different project effort values registered on two project tools, and the related mitigation solution proposals, in the same AI prompt request, I only received a list of risks and not quite appropriate mitigation measures. Reducing the request to the list of risks and iteratively asking for the solution proposals per listed risk, the result was much better. Saving Changes...
LUIZ FREITAS JUNIOREngineer| Tranship transportes maritimos ltdaRio De Janeiro, Rio De Janeiro, Brazil
You must be specific, clear, unambiguous, give clear instructions and examples. Interact with generative AI by refining your query step by step. By following these steps you can produce higher quality results Saving Changes...
Mikaella DarumOperations Effectiveness Manager and Change Manager| RELX Reed Elsevier (Philippines)Quezon City, Philippines
It's amusing as prior to undergoing any formal prompt engineering workshop, I've been "talking" to GenAI like how I'd talk to a child or a friend. Given that I now know formats such as RTF and CREATE, the essence of refining a prompt is indeed like speaking to an extremely intelligent entity who can't recognise the nuances. With this train of thought, I patiently break down my requests to sub-tasks, give it context, until I get the satisfactory response. Saving Changes...
Shea KileyEducator| Massachusetts Department of CorrectionMassachusetts, United States
Using an AI specific to the attempted task is a solid first step toward getting your desired result. Also, taking time to read prompting instructions, or best practices for the specific Ai being used, should not be overlooked when getting started. Saving Changes...
Shea KileyEducator| Massachusetts Department of CorrectionMassachusetts, United States
I have refined prompts by changing and switching out words with their synonym or even a trademarked made-up marketing term. I would suggest folks should try this to check how basic results vary. Saving Changes...
Shea KileyEducator| Massachusetts Department of CorrectionMassachusetts, United States
Jun 21, 2024 11:10 AM
Replying to Sergio Luis Conte
...
It is important to remember this: generative AI is just "predictive text with storoids". Obviously not only text will be the result. BUT the important thing is the answer will just to complete your question (prompt) with the things that have more probability to complete it. You can manage it using some of the parameters like temperature. So, it is very important when creating the prompt to put clear the role, the place where the role works/live/etc, the task the role has to accomplish and the format of the answer. This is an example of R-T-F. You have to eliminate as ambiguity as possible. If not, then hallucinations will happened.
PMI Team Member
